1. UKM Learning and Research Repository
2. Master Thesis / Tesis Master
3. Faculty of Information Science and Technology / Fakulti Teknologi dan Sains Maklumat

Title:	Automated visual inspection for surgical instruments based on generalization of Bag-of-Words model
Authors:	Murtadha Basil Abbas (P56104)
Supervisor:	Anton Satria Prabuwono, Assoc. Prof. Dr.
Keywords:	Image processing -- Digital techniques -- Mathematics Engineering inspection -- Automation Computer vision Universiti Kebangsaan Malaysia -- Dissertations Dissertations, Academic -- Malaysia
Issue Date:	30-Aug-2012
Description:	Upon increasing the pressure of imposing high requirements in industrial sectors, the common and traditional methods of human subjective evaluations are gradually being replaced by the Machine Vision (MV) technique. In fact, this technique is called as Automated Visual Inspection System (AVIS). Along with the last few decades, the AVIS is playing an important role in either side of quality control and/or production process. At current state, AVIS applications are mostly developed based on shape and geometric feature extraction methods. Unfortunately, that approaches have some problems due to image scaling, illumination, noise and other variant issues. However, the main objective of this research is to introduce the spatial local feature to the AVIS development based on the generalization of Bag-of Words model (BoW) and to develop an intelligent image classification algorithm for the AVIS to classify the surgical instruments as a case study in order to measure the efficiency of the proposed method by comparing the classification result with the states-of-art model denoted by standard Scale Invariant Feature Transform (SIFT). Generally, the AVIS methodology covers two parts: the first one is the hardware installation, which consists of webcam, source light and the conveyor belt while the second one is the software techniques that comprising image processing and object recognition. Furthermore, this work is based on three criteria: image acquisition calibration, AVIS environmental conditions and (individual and group) classification technologies. Finally, the obtained results show that the proposed AVIS work can achieve up to 92% and 90% of accuracy rates for individual and group classifications in sequence for both of K-Nearest Neighbour (KNN) and Support Vector Machine (SVM) classification techniques.,Master of Information Technology,Certification of Master's / Doctoral Thesis" is not available"
Pages:	111
Call Number:	TA1637.5.A234 2012 3 tesis
Publisher:	UKM, Bangi
Appears in Collections:	Faculty of Information Science and Technology / Fakulti Teknologi dan Sains Maklumat

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.